The paper introduces OverThink, a slowdown attack that forces reasoning language models (RLMs) to produce many more reasoning tokens while still giving correct answers. By injecting decoy reasoning problems—such as Markov decision processes, language translation, or graphic comprehension—into the model’s context, attackers can dramatically increase token generation (up to 46× on SQuAD and 17× on coding agents). The study evaluates the attack on both proprietary and open-source RLMs across multiple datasets, explores multimodal and coding‑agent variants, and tests several defenses, concluding that defending against OverThink is challenging and that newer RLMs are even more vulnerable due to higher per‑token costs and increased reasoning token usage.
By Abhinav Kumar, Jaechul Roh, Ali Naseh, Marzena Karpinska, Mohit Iyyer, Amir Houmansadr, Eugene Bagdasarian
MemoryArena is a new evaluation gym that benchmarks agent memory in interdependent multi‑session tasks. Unlike prior benchmarks that test memorization or single‑session action in isolation, MemoryArena requires agents to acquire memory while interacting with the environment and then use that memory to guide future decisions across a range of tasks such as web navigation, planning, information search, and formal reasoning. The benchmark reveals that agents excelling on existing long‑context memory tests perform poorly here, highlighting a gap in current memory evaluation methods.
By Zexue He, Yu Wang, Churan Zhi, Yuanzhe Hu, Tzu-Ping Chen, Lang Yin, Ze Chen, Tong Arthur Wu, Siru Ouyang, Zihan Wang, Jiaxin Pei, Julian McAuley, Yejin Choi, Alex Pentland
This scoping review examined 48 studies on generative AI chatbots designed to deliver motivational interviewing (MI). It found that most systems were text‑based and disembodied, with about half incorporating dynamic adaptation, and that safety reporting was inconsistent. While user perceptions were generally positive and many studies reported MI‑consistent interactions, evidence for sustained behavioral or functional change remains limited.
By Runze Hu, Jingqi Kong, Yang Yang, Yihang Yang, Jingyao Liu, Haizhou Tang, Shanghang Zhang, Zheng Liu
The paper presents a traffic sign recognition system that extends YOLOv2 with a branched architecture and geometric feature integration. By adding intermediate prediction layers, the model can terminate inference early for easy cases, reducing computation time, while unsupervised Bayesian segmentation supplies geometric templates to improve classification of visually similar signs. Experiments on a combined GTSDB/GTSRB dataset show that the branched model achieves 0.680 mAP in 0.647 s, and adding geometric verification during inference raises mAP to 0.713.
By Arefeh Rezaei
VeriFuse is a bounded arbitration framework that integrates vision‑language models (VLMs) into vehicle‑infrastructure cooperative 3D perception. Each agent first generates independent detections, then VeriFuse creates a unified candidate pool of geometric proposals and cross‑source hypotheses. A frozen VLM selects among three actions—SELECT, REFINE, or REJECT—to resolve ambiguity and produce final 3D detections, achieving strong AP50/AP70 scores on the DAIR‑V2X dataset while keeping vehicle‑side BEV AP50 drop minimal under delay.
By Hongyi Lin, Yiyao Liu, Qi Kang, Heye Huang, Yang Liu, Haris Koutsopoulos, Jinhua Zhao
The paper proposes a method for learning task-relevant representations in deep reinforcement learning by maximizing rollout total correlation, which captures the correlation among all learned representations and actions across entire trajectories. It introduces two complementary lower bounds—one generative and one discriminative—along with chunk‑wise mini‑batching to improve this objective, and also proposes an intrinsic reward derived from the learned representation to enhance exploration. Experiments on challenging image‑based simulated control tasks demonstrate improved sample efficiency and robustness to white noise and natural video backgrounds compared to leading baselines.
By Bang You, Huaping Liu, Jan Peters, Oleg Arenz
ArenaFlow is a hierarchical credit propagation framework designed to improve reinforcement learning for open-ended agent tasks. It uses tournament-based relative ranking to generate trajectory-level rewards and structured reflective evaluation to identify pivotal success steps, reusable strategy skills, and skill usage attribution. The framework propagates advantages to high-confidence steps and maintains a global skill memory, enabling more targeted optimization and reusable skill priors for future exploration.
By Qiang Zhang, Ruixue Ding, Fanrui Zhang, Xi Chen, Boli Chen, Shihang Wang, Yinfeng Huang, Yi Zheng, Pengjun Xie, Kaipeng Zhang, Jiawei Liu, Zheng-Jun Zha
The paper introduces Omni Demand Understanding (ODU), a benchmark designed to test whether multimodal models can infer a user's underlying demand from complex audio‑visual interactions. ODU requires models to detect the presence of a demand and infer intent using multimodal and conversational context, evaluated across single‑turn and multi‑turn scenarios. The authors built ODU‑Bench through a taxonomy‑guided approach, agentic video generation, and human‑recorded interactions, and found that even top models like Gemini 3.1 Pro recover only 44.7% of key information, with many models exhibiting high false‑trigger rates.
By Qi Chen, Yunfei Chu, Haolin He, Yifan Yang, Zihan Liu, Yuxuan Wang, Ziyang Ma, Ruiyang Xu, Meng Gao, Yinsong Yan, Ling Wang, Hui Wang, Wen Huang, Yiheng Chen, Guanrou Yang, Qiuqiang Kong, Jin Xu, Xie Chen
MIRAGE is a new inference-time framework that enhances large language models by using a Selector to choose effective conceptual perspectives and a Reasoner to solve tasks step-by-step, aggregating multiple perspectives when needed. It is inspired by human cognitive flexibility and is designed to improve performance on complex mathematical, scientific, and logical problems. Experiments on GSM8K, MATH500, MMLU-Pro, and Game-of-24 show that MIRAGE outperforms Chain-of-Thought and diverse prompting ensembles, boosting accuracy with minimal inference overhead.
By Arash Lagzian, Srinivas Anumasa, Dianbo Liu
GestureFAR is a flow‑autoregressive framework that generates natural co‑speech gestures from streaming speech while preserving causality and continuous motion expressiveness. It autoregresses over continuous motion latents using a transformer for audio‑motion context and a flow‑matching head to sample the next latent. A head‑only flow distillation strategy further reduces latency by collapsing multi‑step flow sampling into a single network evaluation, enabling real‑time token‑causal generation with improved quality‑latency trade‑off on the BEAT2 benchmark.
By Pinxin Liu, Haiyang Liu, Jiahao Luo, Junhua Huang, Chunhao Zou, Luchuan Song
SkillIR is a skill-guided framework for agentic image restoration that represents restoration experience as degradation-centered action evidence rather than full tool-use trajectories. It consolidates context-dependent action outcomes into scene-aware restoration skills, guiding one bounded action at a time within a verified residual-state loop. Experiments on synthetic and real-world multi-degradation datasets show that SkillIR improves restoration quality and enables more reliable and effective tool use.
By Jie Shao, Shengkai Hu, Xu Zhang, Beihang Song, Yongcheng Jing, Xu Wu, Jun Wan
ULTRA is a unified framework for autonomous humanoid whole-body locomotion and manipulation that overcomes limitations of prior methods by combining a physics-driven neural retargeting algorithm with a multimodal controller. The retargeting algorithm translates large-scale motion capture data into physically plausible humanoid motions, while the controller learns to handle both dense motion references and sparse task specifications using a range of sensory inputs, from accurate motion-capture states to noisy egocentric vision. In simulation and on a real Unitree G1 humanoid, ULTRA demonstrates improved generalization and robustness, enabling coordinated whole-body behavior from sparse intent without relying on test-time reference motions.
By Xialin He, Sirui Xu, Xinyao Li, Runpei Dong, Liuyu Bian, Yu-Xiong Wang, Liang-Yan Gui
The paper introduces Survival Reinforcement Learning (SRL), an online classification-based method that extends the survival value learning framework to maximize an agent’s dwell time at target goals. SRL addresses limitations of contrastive reinforcement learning (CRL) in long-horizon, goal-conditioned tasks by avoiding the uniformity-tolerance dilemma and reducing undesirable bang-bang control behaviors. Across robotic benchmarks, SRL matches CRL on manipulation tasks and outperforms it by 2x to 8x on stable, long-horizon locomotion tasks, suggesting classification-based approaches are a promising direction for scaling reinforcement learning.
By Franki Nguimatsia-Tiofack, Fabian Schramm, Th\'eotime Le Hellard, Justin Carpentier
arXiv:2511.13254v2 Announce Type: replace
Abstract: Large Language Models (LLMs) have displayed remarkable capabilities across diverse domains, but their training remains resource- and time-intensive...
By Shalini Maiti, Amar Budhiraja, Bhavul Gauri, Gaurav Chaurasia, Anton Protopopov, Alexis Audran-Reiss, Michael Slater, Despoina Magka, Tatiana Shavrina, Roberta Raileanu, Yoram Bachrach
arXiv:2504.06647v4 Announce Type: replace
Abstract: Safety-critical autonomous driving motivates the effective use of prior information. For online vectorized HD map construction, temporal prediction...
By Nan Peng, Xun Zhou, Mingming Wang, Guisong Chen, Wenqi Xu
arXiv:2603.22161v3 Announce Type: replace
Abstract: Metacognition -- assessing the quality of one's own cognitive performance -- guides adaptive behavior across species. Substantial research demonstr...
By Dharshan Kumaran, Nathaniel Daw, Simon Osindero, Petar Veli\v{c}kovi\'c, Viorica Patraucean
arXiv:2609.21293v1 Announce Type: new
Abstract: Autonomous software generation (ASG) aims to turn human requirements into executable applications, but delivering these applications does not necessari...
By Xiuhui Zhang, Yi Chen, Shusheng Xu, Fan Li, Huan Wang, Tongkai Yang, Binhang Yuan
arXiv:2609.21486v1 Announce Type: new
Abstract: Multimodal trajectory prediction improves behavioral coverage in end-to-end autonomous driving, but existing methods remain limited by sparse scene rep...
By Jiaxing Chen, Hengduo Zou, YuKai Qin, Yiren Zhao, Lidong Yu, Bolin Gao
The paper proposes that Learning-to-Optimize (L2O) should serve as a missing architectural layer in AI-native communication networks, bridging optimisation and AI intelligence. It redefines optimisation algorithms as offline knowledge generators that produce supervisory data for neural surrogate models, enabling low‑latency inference in dynamic environments. A four‑stage workflow—optimisation, knowledge generation, surrogate learning, and runtime inference—is introduced, and demonstrated on an NR‑V2X relay‑selection problem where a Graph Neural Network learns near‑optimal decisions from MILP solutions.
By Giambattista Amati, Federica Mangiatordi, Pierpaolo Salvo, Emiliano Pallotti, Simone Angelini
arXiv:2609.21841v1 Announce Type: new
Abstract: Frontier language models now produce professional deliverables that expert graders judge to match human work on a substantial share of economically val...
By Abbas Raza Ali, Muhammad Ajmal Siddiqui, Moona Zahid